Japan checks on thorax is not relevant

[Ipuch]

• Nishinaka et al 2008 (GH)
• Matsuki et. al. 2011 (ST)
• Mastuki et al. 2012 (GH
• Matsuki et al. 2014 (SC)
• Kijima et al. 2015 (GH, ST)
• Kim et al. 2017 (ST)
• Sugi et al. 2021 (ST, GH)

Thorax is a global frame but is not used as the fluoroscope is centered on the scapula frame.
Only if the joint does not include the thorax.

We ask ourselves if we can include scapulothoracic and sternoclavicular in in Matsuki datasets dans japan datasets

• @fmoissenet send an email to the authors
• authors responses:
  • Nishinaka et al 2008 (GH)
  • Matsuki et. al. 2011 (ST)
  • Mastuki et al. 2012 (GH
  • Matsuki et al. 2014 (SC)
  • Kijima et al. 2015 (GH, ST)
  • Kim et al. 2017 (ST)
  • Sugi et al. 2021 (ST, GH)

We may rely on a check on how thoracohumeral angle has been computed. if it's controlled by the operator or a projection between two axes, then no correction can be applied on thorax' frame and we can just ignore the data in its frame and utimately scapulothoracic, sternoclavicular data should be thrown out.

Below the answer from S. Banks concerning the general process applied to these papers for the "thorax" frame:
"(1) Because we did not have the ability to track the thorax within the radiographic frame, the protocols were very specific to keep the thorax aligned vertically in the room, and to minimize any angular deviation of the thorax during the recorded motions. (2) We generally found the 3-1-2 (Z-X-Y, abduction/flexion/axial rotation) Cardan decomposition to work well. I do recall we played with the ISB Euler sequence and others over time, but the specifics should be stated explicitly in each paper. When the 3-1-2 sequence was used, it would have been used to describe the scapula relative to the room-fixed coordinate system, and for describing the humerus relative to the scapula. Where the humerus elevation angle was used as an independent description, that most often would be a projection angle relative to the room-fixed axes. "

Still waiting for answers from Nishinaka and Kim ... but I would assume that the same approach as been used, i.e. thorax is global.